Miriah Meyer is an associate professor in the School of Computing at the University of Utah and a faculty member in the Scientific Computing and Imaging Institute. She co-directs the Visualization Design Lab, which focuses on the design of visualization systems for helping people make sense of complex data, and on the development of methods for helping visualization designers make sense of the world. She obtained her bachelors degree in astronomy and astrophysics at Penn State University, and earned a PhD in computer science from the University of Utah. Prior to joining the faculty at Utah Miriah was a postdoctoral research fellow at Harvard University and a visiting scientist at the Broad Institute of MIT and Harvard.
Miriah is the recipient of a NSF CAREER grant, a Microsoft Research Faculty Fellowship, and a NSF/CRA Computing Innovation Fellow award. She was named a University of Utah Distinguished Alumni, both a TED Fellow and a PopTech Science Fellow, and included on MIT Technology Review’s TR35 list of the top young innovators. She was also awarded an AAAS Mass Media Fellowship that landed her a stint as a science writer for the Chicago Tribune.
The PolicyViz Podcast is now on YouTube! More past episodes being added soon. Subscribe to the channel to get more podcasts, tutorials, and more data visualization content!
Support the Show
This show is completely listener-supported. There are no ads on the show notes page or in the audio. If you would like to financially support the show, please check out my Patreon page, where just for a few bucks a month, you can get a sneak peek at guests, grab stickers, or even a podcast mug. Your support helps me cover audio editing services, transcription services, and more. You can also support the show by sharing it with others and reviewing it on iTunes or your favorite podcast provider.
Welcome to the 199th episode of the PolicyViz podcast. I’m your host, Jon Schwabish. I hope you are well and healthy and safe, and I hope you are going to continue to listen to the podcast as we near the end of this season. Rounding out in just a couple of weeks with the 200th episode, I have something very special planned, I hope you will tune in. But I hope you will also listen to this week’s episode of the show, because I have a very special guest with me, I chat with Miriah Meyer who is an associate professor in the School of Computing at the University of Utah. Miriah and I got to know each other earlier in the year a little bit better after having run into each other a couple of times in the past, but we got to work together on a panel helping a US federal government agency improve how they were visualizing their data. So we had a lot of fun discussions about techniques and approaches and processes and workflows and everything that an organization or an agency would need to have a better data communication setup. In today’s interview, in my discussion with Miriah, we talk about her research. She’s done a lot of research on participatory design workshops, and we talk about how data visualization instructors can utilize some of those methods. We talk about data visualization tools – she did some work a couple of years ago on a tool called Origraph, so we talked about her work on that project and the data visualization tool field sort of generally. And then we also talk about the current state of data visualization research, we talk about what we think or what she thinks the field should move in what directions or different dimensions. There’s also a special reveal coming up in this episode right towards the end of our interview, so make sure you listen to the whole thing. It’s a very special reveal, I’m not going to tell you what it is right now, you’re going to have to listen to the entire episode, and you will, I think, find it to be quite some interesting news, something that I think we will all be looking forward to see what happens over the next couple of years. So I’m not going to delay any longer. I’m going to get right into my conversation with Miriah Meyer. I hope you will enjoy this week’s episode of the PolicyViz podcast.
Jon Schwabish: Hi Miriah. Great to see you. How are you? Welcome to the show.
Miriah Meyer: I’m great. How are you, Jon? Thanks for having me here.
JS: Of course. I’m doing well. It’s great to see you. Now we actually get to see each other. It’s been a few months since we actually chatted last time. We were doing this very cool advisory board panel thing.
MM: Yes, where we basically just got to talk at length about the things we feel very strongly about when it comes to visualization recommendations, it’s lovely.
JS: Yeah, it was. It was just like a free form discussion about everything we love and hate about DataViz. That was pretty good. So we’ve got a lot I think to talk about, all of your awesome work, and maybe some thoughts you have on the future. I thought maybe we’d start with the work you’ve done in the past on participatory design workshops. So you’ve written I think a couple of papers on that. And so, I was hoping maybe you could just talk about that work a little bit, and then maybe I’ll just make a personal, like, how do you implement those ideas into your own teaching.
MM: Yeah. So I’m glad you asked about those. So a number of years ago, one of my students, Ethan Kerzner, he was working on a project with some neuroscientists, and that’s sort of the basis of the work that we do in my lab. It’s about working with people in the world and designing tools for them and using those experiences as ways for us as researchers to ask questions. And so, my student, Ethan was working with this group of neuroscientists, and he was doing the sort of initial, you know, before we design, trying to understand what it is that they’re having challenges with, what are visualization opportunities for us, and he would go and interview one lab member, and they’d say, this is the really important thing; and then he’d go and interview another lab member, and they’d say, no, no, no, it’s this thing; and then the PI for the lab would be like, really, none of those things are important. And so he was getting really frustrated. And so, about that time, he saw a paper that was published by a group in London, they were working with these energy modelers, very different group, but they were talking about these workshops that they had been running with this group to better understand to sort of build consensus and understand what the opportunities were.
So, Ethan decided to try this. So he designed and ran one of these, at the time, we were calling them creativity workshops, with our neuroscience colleagues, and it was just incredibly successful. It’s a participatory style workshop where you bring people together. The ones we do are based upon creativity theory, so it’s about how do you get groups of people to be creative together and to brainstorm, but how do we do that in a structured way to get towards very tangible visualization opportunities. And so, this was really – it was great. Not only did we learn a lot, and Ethan was able to – he developed a really nice tool that he deployed to the lab based upon this. But all kinds of other, like, really squishy things came out of it, like, there was this one member of the lab who Ethan just couldn’t get any time with. And after this workshop, this person was then willing to meet with us very regularly. We heard back from the PI, that as a lab, they had made a lot of progress on their own sorts of needs for the next couple of months through these conversations. So there’s all this agency within the group that these workshops developed, and we also were able to build trust among these folks in the neuroscientist. So it was great.
And so, after that, Ethan went and talked with the London group about their own experiences; and then we spent the next two and a half years, reflecting, as a group, on our experiences of running many of these workshops, and we were able to formalize that into a framework, a sort of structured framework for how you design these workshops, run them, and then analyze them. And so, after we did that, we started using these workshops in every single project that we do, and I really think it saves months and months of time that otherwise you would need to spend talking to one person and then another. And so, it’s a way to bring people together and get a consensus on what some opportunities are for visualization design. And I think there’s probably many people who listen to your podcast like practitioners and other scholars who do similar kinds of workshops. I think it’s a very common and productive method to bring people together, but I think what’s useful about the workshops that we developed is that they are super structured, in that we have certain activities that we like to do. There’s a sort of ebb and flow of making people very much diverged in their ideas, but then bringing them back together and doing that over and over, and how that can lead towards some sort of consensus building on critical ideas. So that’s, I guess, a long winded answer to your question, but they’re awesome, and they’re very much tailored for trying to find out visualization needs.
JS: So I want to ask you, because the way the workshop was developed, was through this sort of qualitative approach, so talking to people having these conversations, and I know you do other research that has more of a quantitative bent. And so, I’m curious about your take on the blending of the two, and specific obviously to Viz research, but do you have – I don’t really know, I don’t honestly really know what the question is, but, do you feel like it’s important to merge the qualitative and quantitative? Should they be separate? I assume you’re trained as a quantitative researcher, not as a qualitative researcher. So what has that sort of switch in your, or not switch, but added skill set to your approach been like?
MM: Yeah, I love that question. And in fact, in thinking about these workshops, well, let me – so yeah, my own background, my undergrad degree is in astrophysics, and then my PhD is in computer science. My PhD work was very quantitative in nature, but if you look at the kind of work I’ve been doing the last couple of years, it’s almost exclusively qualitative. So there was a real inflection point. And I think that our developing of the sort of workshop framework is sort of a nice case study of how that switch went. So again, my student, Ethan, that was leading this work, he was also very – he’s very quantitative person; and when he had to tackle the problem of trying to analyze all the stuff that we – all the sort of data and evidence that we collected during the workshop, he started by making these spreadsheets and counting things, like how many insights do people have, and can we rank those on a variety, can we assign numbers to them for a variety of attributes. And at the end, he kind of threw all that out, because it’s like, it doesn’t matter how many times someone said this word, like, that doesn’t matter. But what really mattered more was his own sort of reflection on what he learned, sort of, observing and facilitating this group of people that he was going to work with.
And so, that was like a really difficult shift for him to let go of, and I have to say, like, I sort of went along with him in that journey. But I think sort of more generally, my own feeling is, when we’re looking at visualization in the world, I just don’t think putting numbers on things really tells us the kinds of things that matter to someone who’s going off and designing for the messy complexity of people in the world and relationships and all that stuff that gets entangled together. And so, my own work has very much shifted towards qualitative work. And, in fact, a whole project led by another student, that is all about – it’s a longitudinal study where we’re doing lots of interviews along the way, and even there, we sort of abandoned the idea of open coding, which is, I think, the sort of quantitative person’s way of approaching qualitative work, like, let’s put some codes and then count them and count how many codes match between this coder and this coder. At the end, I just feel like, it’s really hard for us sometimes to value what we just intuitively know and learn. But that is where I think the knowledge resides, it resides in my interpretation and my students’ interpretation and then us talking. And so that’s actually led to a whole other stream of research I’ve gotten interested in, which is how do you document this sort of reflective collaborative process. So how do you document it? How do you provide evidence? How do you convince people that the outcomes you came to are reasonable and plausible? And then also, how do you even allow people to understand how you got there? And so that I think is a whole another sort of interesting, deep bit of work, you know, we’re starting to think about, but it’s very much in support of people going off and doing messy things in the world, but valuing what we as individuals learn from that.
JS: Right. So looking at the DataViz research field sort of more broadly, do you think that the field has that right balance or not? I can tell you how I feel about the econ field, but not about DataViz. So what’s the balance like in DataViz?
MM: It’s like you’re trying to poke me.
JS: [inaudible 00:11:52] cause some trouble in the early summertime here.
MM: Well, I have to say, no, we don’t have the right balance. And I think on the whole the visualization research community knows that, like, every year, there is a recognition of we need more qualitative work, we need more examples of qualitative work. I’ve been in the community reviewing for many years. I’ve [inaudible 00:12:18] in the last couple of years, and sort of seeing the kinds of standards that are applied to qualitative work is a little bit heartbreaking, because it’s really hard. And it’s also a bunch of methods and skills that your typical computer science like graduate or undergraduate student doesn’t have any training at.
JS: At all, yeah.
MM: And so many of us are sort of winging it, and making it up and doing our best, and it’s really hard when you position that kind of work against what has become a very well established, quantitative bent in our community, like controlled studies and statistical analysis. There’s people in our community now who are clearly experts and leading in that, but that is so well established, and you can name the kinds of statistical tests to do, you look at qualitative work, and really, it’s so individual, and it’s so specific to the project and to the needs and to the skills. And that’s okay, that it makes it really hard to sort of, you can’t have a checklist of things and say, this is good work, because it means a checklist, it doesn’t work for qualitative in the same way. I think it works for quantitative. So as a community, I think we’re really struggling with that, and I feel like we’ve gotten really good at the sort of more positivist, objective, loving, controlled science side of things. The community has, I think, really come a long way there. But because of that, now that it’s making it in some ways, I think, more difficult for a design oriented and qualitative work to get the same kind of recognition and acceptance, because they’re just fundamentally two different approaches to research, and yet, as a community, we tend to just apply one sort of epistemology or perspective and how to value it.
JS: Yeah, I know…
MM: Okay, sorry [inaudible 00:14:09]
JS: Well, I was going to say, I think the same is generally true for economics. There’s been, over the past couple of decades, the growth of the behavioral economics field, which is more qualitative, but it’s also sort of a slice or a sub, it’s kind of a sub sector of economics, where, I think, to your point, I would guess that CS grad students and economics grad students have the same thing that there’s little to no training on qualitative methods. Whereas if you go to a sociology department, it’s – not that I’ve ever – I’ve only gone to sociology webinars to be clear, sociology webinars to get free lunch, that was the plan in graduate schools working your free lunch. But sociology graduate students, the ones that I know, they’d mostly qualitative skills, but they also had some quantitative skills, like, they knew how to clean data and run a basic regression. I mean, if you throw [inaudible 00:15:05] higher level econometrics at it, that’s not something that they would know. But they had all these qualitative skill sets that like, I didn’t have to read about it. So I wonder, you’re a faculty member, how is the evolution of the field, how does it move forward, given the current sort of syllabus, as it were, for graduate students – is that where things start to change, or does this start to change with the more senior level faculty members, researchers who start to embrace these methods, and then it trickles into their grad programs?
MM: That is such an interesting question. Unfortunately, it’s probably not what I thought about deeply…
JS: Yeah, [inaudible 00:15:46]
MM: Yeah-no, you know, one thing that strikes me is the social science department that I know of, for example, here at the University of Utah. They have methods classes that span quant and qual. They have methods that teach them, there are different epistemologies out, there are different views on what knowledge is and how we come to know it. And they teach these things in the framing of there isn’t just a singular worldview about what truth and knowledge is, there are several. And you make decisions about which one is the best fit for the type of inquiry that you’re doing. And you look at computer science programs, and it sounds like probably similar for economics, and we just don’t have that training. And I think the problem then is that there is an approach to how we think about knowledge and truth. We don’t even know it. We don’t even recognize that we have a singular perspective. We don’t even know that it’s a perspective.
MM: And so, I look to – one of the places that I find a lot of inspiration in are iSchools, and there’s a number of very technical iSchools now that I just find the work coming out of there so exciting. And you look at their curriculums and the students have to take classes that teach you about different philosophies and these kinds of things. And that’s the kind of thing that I would love to see sort of coming into more computer science programs. I think the fact that as a field, we suddenly realize, hey, ethics is like a thing, and there’s like a problem here. And we don’t really know how to handle that, and that’s because, right, we don’t have the training yet. And so, a lot of programs, including my own department, we’ve been working on building ethics curriculum into our degree programs. And I think that that’s a step, but I think that ethics also comes from recognizing that there’s different types of thinking besides math and logic. And so, where am I going with all this – I think in computer science, I think getting training beyond basic programming and object oriented programming and operating systems, there’s all these other things that I think are increasingly important. They have been important for a long time, but only recently I think has, like my community, my CS community recognize that, oh yeah, maybe we need to do something different because we seem to be causing harm, and we don’t know how to solve it.
MM: So I think it’s kind of a great time to sort of make these kinds of changes, and I think there’s, as you were sort of implying, there’s a lot of opportunities to train this next generation definitely, and to understand that there’s many different ways to think about the world. And I think if the younger kids all do it, then the senior people have no choice but to go along with it.
JS: Right. Yeah. I mean, it is interesting, because it’s always a tradeoff. So if you have to take X number of classes, and we’re going to say you have to take ethics, and you have to take philosophy of thought, well, there’s two classes that you have to take out somewhere else. But I’ve seen lots of – information schools is a good example – I’ve also seen a lot of public policy schools trying some different ways to provide, well, in DC, it tends to be very professional development in the schools that I work with. So it’s like, how to write for a policy audience, how to do GIS software, and so it’s not like a full semester class, but it gives you that, at least, that introduction to those skills that are not typical in these different fields or different departments.
MM: Totally, yeah. It’s funny, like I struggle a lot in our own department, exactly what you’re saying, it’s like, well, these classes seem important, but what about all this other important stuff.
MM: And at least for computer science, it seems like there’s this really, I think, exciting trend of recognizing that computer science is broad, and there may not be just one flavor of what CS means, and explore, like, the person who’s going to be doing a lot of frontend development work, may not need to know about networking or how a compiler works. Right? Like, at some point, we can’t know it all. And I suspect if you look at fields that perhaps are much more fluent and deeply embedded in methods, a lot of them are sort of being asked now to be computational as well. And so, I know that they’re struggling in similar ways to.
JS: Yeah, I bet that’s true. I want to turn back to another strand of your research, which is on data visualization tools. So you did this paper, I think it was fairly recently, right, on Origraph. And so, I was wondering if you could tell folks a little bit about that project, and then we can talk sort of generally, maybe about DataViz tools and where you think tools are going over the next several years.
MM: Yes. So that project, Origraph, this was a project that was led by another student of mine, Alex Bigelow, and the idea behind it was really about graph wrangling. And so, what we wanted to do was think about graphs or networks as a representation are a really interesting representation for thinking about relationships between things in your data. And a lot of the things that we think of as networks or as a graph aren’t necessarily physically connected things in the world. So it’s more of an abstract sort of thing. We think about a social network, like, I’m not physically connected to you, Jon, but because I follow you on Twitter, I have a connection to you. So we abstract that into this sort of graph relationship. Anyway, so we got really interested in how people use graphs as a sort of thinking tool. And we started thinking about like, well, what would it mean to allow people to change that graph model; what would it mean to allow people to define on the fly, what it means for someone to have a connection, like, who am I connected to that also does data visualization and lives on the East Coast, and now I have a connection to you. And we looked around, like, there’s a lot of data wrangling for tabular data and things like that, there’s tons of exciting and great tools out there for that, but graphs still hadn’t been done.
So the whole idea of Origraph was to provide a tool that allowed people to define a model for a graph, and then be able to visualize it. And I’m excited that of all the tools I’ve created that I’ve been part of that you picked this one, one of the reasons I like that is because to me that also sort of embodies what I see as sort of a major future thrust of visualization, which is thinking less about the specific visual encoding challenges, and more about the visualization as a tool. And so, I think the heyday of the Viz community, at least, just the research community is all I can speak for, coming up with like, really exciting new techniques that actually like matter, is largely gone. But instead, it’s about experimenting and thinking about what visual representations and interactive representations allow us to do. And, of course, with data science being so important, I think visualizations are a really critical way of helping people reason about data. So that’s one of the things I see as being increasingly important for visualization is less of a focus on the specifics of how I encoded something, and more about thoughtful design work that gives people access to data and opportunities to explore it. And this tool Origraph was, I think, an example of that.
JS: It’s really interesting, because of the projects that we’ve talked about, like the through line of your work seems to be how do we bring people together to solve problems or be creative within a group in the design workshops, like, in person, and then through a data visualization tool where it’s being creative or connecting people within the tool itself.
MM: Yeah, I love that perspective. I have to say, like, my work is less and less about tools, and more and more about what we learn through the act of either designing visualizations or people using them. And so, I think it’s very much as sort of the human side of data science and visualization is – that is our tool, our probe, our technique for exploring that.
JS: Right. So you have, I’m sure, a lot of projects starting up ongoing. Do you have anything you want to talk about briefly? And then I know you have a big move coming up, starting up a whole new project so we can talk about all that stuff. So maybe we can talk about the big move and then what you hope to accomplish after you’ve packed up all the stuff in your house?
MM: Well, since you asked Jon, yes, I do have a big mirror.
MM: Yeah, I’m really, really excited. I’ve accepted a new position at Linköping University, which is in Sweden. There’s a big visualization group that’s been there for quite some time now, that’s headed up by Anders Ynnerman, and they do some just amazing work, particularly with public outreach. They have a big museum space, and think about how visualization and data can be used in these sort of communicative outreach settings. So yeah, so I’m really excited to go and try some new things and build some new collaborations. And so, yeah, so I’m going to be moving there this fall. And one of the things I’m really excited about moving to particularly, Scandinavia, is the sort of – it’s kind of the birthplace of participatory design and thinking a lot about democratization of technology and data and all these things. And so, I’m excited to sort of be embedded in a different way of thinking about how we make visualization and data more inclusive and more diverse. So that is a theme of mine, I think going forward, both from me thinking about it theoretically, from an epistemological perspective, I’ve been doing a lot of work with applying feminist theory to visualization recently; but then also sort of questioning like, how we – who we’re designing tools for, and is designing tools always the thing we should be doing.
And I have a project with putting air quality sensors in the homes of people with asthmatic kids, this is a project led by another student of mine, Jimmy Moore. And after having spent several years with these families and interviewing them, and ultimately, we had hoped to design a visual analytics tool to help them explore their data, our end conclusion is actually, the most effective thing that we did was to bring an analyst into their homes and let them ask questions, and in real time, show them stuff. And that’s really led us the question, would it have been worth all the resources and hours of designing the tool that maybe they would use, and maybe instead, we should think about systems, how could we scale up that idea of having access to train data analysts in some sort of, like a data clinic – what would it look like to have data clinics and to build a volunteer corps of people like us that would be interested in working with others, and what are the incentives there? So those are the kinds of things that I’m excited to work on, and I feel like a place like Sweden and Europe, in general, there’s a lot of history of those kinds of ideas coming out of those countries.
JS: Right. That’s very exciting. Congrats [inaudible 00:27:50]. That’s exciting. I want to ask, before we round up, you had mentioned that there’s a big museum and planetarium there. So do you see yourself, and it sounds like you do see yourself working with the folks who are in those museum spaces, because I feel like DataViz in museum spaces is like an underappreciated part of the world of database, like, here in DC, when you go to the Air and Space Museum in downtown DC, they have these little posts that represent with a little like sphere on top for each planet and they’re sized and spaced out accordingly. So Jupiter’s like, whatever it is, five blocks away or something like that, and I just feel like that’s such a great way to visualize data. And I don’t think it’s a field really talked about that much. So I’m curious, do you see yourself working with that team?
MM: I would love to. I think one of the things that I’m excited about this group I’m going to be joining as they also have learning science – they have a learning science faculty there who are part of the people that think about how people learn in the museum setting. And so, it’s sort of bringing together a design group and stuff. And yeah, and I think as a community, the Viz community, some people have tried to touch on that, but we get so wrapped up and represent the sort of accurate, I’d say, in air quotes, accurate representation of values that we forget about the sort of embodied nature that, like what you’re describing this sort of very visceral reaction of [inaudible 00:29:23] like Jupiter’s five blocks away, oh my goodness. And I think there’s so much that we can learn about how people – the experience of representations and what that might mean for how we think about visualization. There’s this whole emerging thing with data physicalization, and what does that bring. And it’s easy to say, oh data physicalization is great for teaching people about Viz or for getting people excited about their personal data, but why aren’t we doing data physicalization for bioinformaticians, why don’t they get to play and enjoy too. So I think like you’re saying, there’s a lot of crossover, I would think from what we as a community can learn from museum settings and other sorts of engagement approaches, that I think would be super exciting to apply to our working with scientists, because I think we’re kidding ourselves, if we think that a scientist is just this dry person who looks dryly and very, like, they’re like a [inaudible 00:30:26] they’re just this perceptual machine looking at their data. But no, these people have hopes and dreams and feelings too. And I want them to have to look five blocks away and be like, wow.
JS: Yeah, right. Exactly. Yeah, I think that that is exciting. I mean, I definitely am excited to see what comes out of your new position, new experience, because I feel like what we’ve been talking about this through line of this getting people together and sort of this crosscutting thing that seems to go through all of your research, really can just blossom and find some really interesting patterns and combinations and teams, and where you have access to a whole other team that’s doing this sort of interesting type of work that maybe we don’t interact with enough.
MM: No pressure, right?
JS: No pressure at all, and I will just keep refreshing your website every month, like, starting in September.
MM: Thank you Jon.
JS: All right, thanks so much for coming on the show. This has been great. Congrats again on the new position. I’m sure everybody who’s listening is going to be excited too.
MM: Thanks, Jon. This was such a lovely conversation.
JS: Great to see you again. All right, thanks so much.
Thanks to everyone for tuning in to this week’s episode of the podcast. I hope you enjoyed listening to that conversation with Miriah Meyer. I hope you’ll go check out her work. I’ve put all the links to the things that we’ve talked about including her research in the show notes for this week’s episode of the show.
So we get down to the last episode of the PolicyViz podcast for this season in two weeks for episode number 200. Make sure you come back, check it out, something very special coming your way. So until next time, this has been the PolicyViz podcast. Thanks so much for listening.
A number of people help bring you the PolicyViz podcast. Music is provided by the NRIs, audio editing is provided by Ken Skaggs and each episode is transcribed by Jenny Transcription Services. If you’d to help support the podcast, please share it and review it on iTunes, Stitcher, Spotify or wherever you get your podcasts. The PolicyViz podcast is ad free and supported by listeners. If you’d like to help support the show financially, please visit our Patreon page at patreon.com/policyviz.